Paper: Discriminant Ranking for Efficient Treebanking

ACL ID C10-2166
Title Discriminant Ranking for Efficient Treebanking
Venue International Conference on Computational Linguistics
Session Poster Session
Year 2010

Treebank annotation is a labor-intensive and time-consuming task. In this paper, we show that a simple statistical ranking model can significantly improve treebank- ing efficiency by prompting human an- notators, well-trained in disambiguation tasks for treebanking but not necessarily grammar experts, to the most relevant lin- guistic disambiguation decisions. Experi- ments were carried out to evaluate the im- pact of such techniques on annotation ef- ficiency and quality. The detailed analysis of outputs from the ranking model shows strong correlation to the human annotator behavior. When integrated into the tree- banking environment, the model brings a significant annotation speed-up with im- proved inter-annotator agreement.†